Skip to content

Education Demo ​

GhostAI for universities, colleges, and K-12 systems. This demo follows Dr. Maria Santos, VP of Information Technology at Pacific State University, as she uses AI for student advising, research support, institutional knowledge management, and academic planning — all with student records, financial aid data, and researcher information automatically protected under FERPA and COPPA requirements.


Pre-Demo Checklist ​

  • [ ] GhostAI instance running
  • [ ] Industry vertical set to Education
  • [ ] FERPA compliance profile active
  • [ ] Memory system enabled for institutional knowledge

Demo Script ​

Chapter 1: FERPA Student Record Protection ​

Duration: 3 minutes Story: Dr. Santos drafts an academic advising summary that includes a student's name, ID, date of birth, GPA, financial aid details, and parent contact information — all FERPA-protected education records.

Input:

Help me draft an academic advising summary for this student:

Student Emily Watson, Student ID PSU-2847103, DOB 05/22/2003, declared Computer Science major fall 2024. GPA: 3.67. Financial aid: Pell Grant $7,395, institutional scholarship $12,000. Parent: Robert Watson, phone (503) 555-0134. Advisor: Prof. Angela Martinez.

Expected Output:

  • FERPA data detected: student name, student ID, DOB, GPA, financial aid amounts, parent name and phone
  • Academic advising summary with student identifiers tokenized
  • Academic performance and major declaration details preserved

Talking Points:

"Student name, student ID, date of birth, GPA, financial aid, parent contact — every one of these is a FERPA-protected education record. GhostAI detected all of them before the AI model saw a single character. The advising summary is still useful — academic standing, major declaration, aid package — but the student's identity stays inside your institution's systems."

Chapter 2: Research Assistance ​

Duration: 3 minutes Story: Dr. Santos helps structure a research abstract. The prompt contains the PI's name, IRB protocol number, research methodology, and study details that need to remain within the institution.

Input:

Help me structure a research abstract for a paper on machine learning applications in educational assessment. The PI is Dr. Carlos Jimenez in our College of Education, IRB protocol #PSU-2024-0847. The study analyzes 5 years of anonymized student performance data to predict at-risk students using ensemble methods. We need the abstract to follow ACM formatting guidelines.

Expected Output:

  • PII detected: PI name, IRB protocol number
  • Structured abstract following ACM formatting: motivation, methods, results framework, conclusions
  • Research methodology preserved; personal identifiers tokenized

Talking Points:

"The PI's name and IRB protocol number are protected, but the research methodology — ensemble methods, five-year longitudinal data, at-risk prediction — stays intact because that's the substance of the abstract. GhostAI understands the difference between who is doing the research and what the research is about."

Chapter 3: Institutional Knowledge ​

Duration: 3 minutes Story: Dr. Santos stores university-wide operational knowledge — academic calendar system, IRB requirements, FERPA training policy, and platform details — into GhostAI's memory, then verifies that a faculty query about IRB requirements is answered correctly.

Input (first message):

Remember this about Pacific State University: We follow the quarter system with 10-week terms. Our IRB requires CITI certification for all researchers conducting human subjects research. FERPA training is mandatory annually for all faculty and staff who access student records. Our LMS is Canvas and we use Banner for student information.

Input (new conversation):

A new faculty member in the CS department wants to conduct a study involving student survey data. What are our IRB requirements and what certifications do they need?

Expected Output:

  • First message: Memory stored, institutional knowledge confirmed
  • Second message: Correct recall of CITI certification, IRB protocol requirements, FERPA training mandate
  • Memory Manager shows stored institutional knowledge

Talking Points:

"A new faculty member asks about IRB requirements and gets an immediate, accurate answer: CITI certification, IRB protocol, FERPA annual training. That's institutional knowledge that used to live in a policy PDF nobody reads. GhostAI makes it accessible in the flow of work — and it's always up to date because Dr. Santos can update the memory any time policy changes."

Chapter 4: Student Advising ​

Duration: 3 minutes Story: Dr. Santos uses AI to draft an academic plan for a Computer Science student with specific credit requirements, course prerequisites, and scheduling constraints across three quarters.

Input:

Help me draft an academic plan for a Computer Science student who needs to complete 45 credits in 3 quarters. Required courses: CS 301 Algorithms (5 cr, offered fall and spring), CS 340 Databases (5 cr, offered all quarters), CS 360 Software Engineering (5 cr, offered winter and spring), CS 370 Operating Systems (5 cr, prerequisite: CS 301, offered winter), CS 410 Capstone (5 cr, prerequisite: CS 360, offered spring only). Plus 20 credits of CS electives. Balance the load across quarters and respect prerequisites.

Expected Output:

  • Quarter-by-quarter academic plan with prerequisite chains respected
  • Balanced credit load: 15 credits per quarter
  • CS 301 placed before CS 370; CS 360 placed before CS 410
  • Elective distribution across all three quarters

Talking Points:

"Prerequisite chains, course availability, credit balancing — this is the combinatorial problem that eats advising hours. GhostAI mapped the dependency graph (CS 301 before OS, Software Engineering before Capstone), slotted courses into their available quarters, and distributed electives evenly. An advisor would spend 30 minutes on this. GhostAI did it in seconds — and no student data left the institution."


SOC 2 Type II Certified | BAA Available | HIPAA-Ready